Secondary Logo

Journal Logo

The Nicolas Andry Award: Modulation of Distraction Osteogenesis in the Aged Rat by Fibroblast Growth Factor

Aronson, James

Clinical Orthopaedics and Related Research: August 2004 - Volume 425 - Issue - p 264-283
doi: 10.1097/01.blo.0000138186.53426.f9
SECTION III: REGULAR AND SPECIAL FEATURES
Free
SDC

This study explores a series of hypotheses related to modulation of bone formation using the distraction model. The tibial lengthening model was scaled down from dog to rat to use immunohistochemistry, in situ hybridization, and reverse transcription-polymerase chain reaction to evaluate cellular events during in vivo bone formation. Different delivery systems such as oral, intragastric, intravenous, subcutaneous, and local diffusion by either extraperiosteal or intramedullary routes, were developed and standardized. Systemic modulators, including diet (total enteral nutrition, calcium, phosphate, soy, whey, casein, lead, and alcohol) and hormones (estrogen, testosterone, growth hormone, and gonadectomy), were tested. To investigate the effects of aging on bone formation, rats of different age groups had tibial lengthening. The aging effect could be distinguished by a reproducible deficiency of endosteal bone formation, consistent with similar deficits in older adult patients having distraction osteogenesis or in patients with senile osteoporosis. Expression of endogenous fibroblast growth factor-2 at the cellular level during the coupled osteogenesis and angiogenesis in young rats was dramatically diminished in old rats. Exogenous fibroblast growth factor-2 reversed the endosteal deficits found in old rats having distraction osteogenesis.

From the University of Arkansas for Medical Sciences, Little Rock, AR.

This work was supported by grants from: NIH RO3 AR 44987-01 (9/97-8/00), NIH RO1 AA12223 (5/99-4/03), USDA #9402890, Brooks Medical Research Fund, Arthur H. Huene Memorial Award (1995), Arkansas Children’s Hospital and the University of Arkansas for Medical Sciences. Recombinant human fibroblast growth factor (rhFGF-2) was generously supplied by Scios and Orquest.

The author certifies that he has no commercial associations (consultancies, stock ownership, equity interest, patent/licensing arrangements) that might pose a conflict of interest in connection with the submitted article.

The author certifies that his institution has approved the animal protocol for this investigation and that all experimentation was conducted with ethical principles of research.

Correspondence to: James Aronson, MD, 800 Marshall Street, Little Rock, AR 72202. Phone: 501-364-1468; Fax: 501-364-1522; E-mail: Aronsonjames@uams.edu.

Bone regeneration by gradual, mechanical distraction of a low-energy osteotomy, termed distraction osteogenesis, is a labor and time-intensive surgical method that has expanded clinical treatment of local bone deficiencies.7 The ring external fixation technique was pioneered by Ilizarov.38–40 Bone lengthening or bone transportation with this method has been used to treat children and adults for congenital hypoplasia, posttraumatic bone deficiency, growth arrest, infection, nonunion, and tumors.3,5,6,8,14,16,27,54,55 Distraction osteogenesis occurs by primarily intramembranous ossification, according to the research of Ilizarov.38–40 New bone is added at a rate of 1 cm/month of external fixation in children.5,27,31 The healing process (bridging and mineralization) is significantly delayed in older individuals.5,27,31

The canine tibial lengthening model, reported by Ilizarov, was reproduced confirming his histologic findings of primarily intramembranous or direct bone formation in uniform microcolumns parallel to the distraction force.2,6,9,12 Various imaging techniques were used to measure different biologic processes observed during distraction osteogenesis. Technetium scintigraphy showed that massive increases in regional blood flow (5–7 times normal for 4–5 weeks and as much as 2 times normal for 4 months) followed the typical injury-response pattern seen with fracture healing; these increases were not prolonged by extending the distraction process.4 Quantitative computed tomography measured mineralization across the distraction gap with greater sensitivity than plain radiography, ultrasound, or dual photon absorptiometry.2,5,6,17,30,67 Quantitative computed tomography when combined with finite element analysis predicted the biomechanical properties of the new bone bridge and could be used clinically to follow the weakest link: the central fibrous interzone by calculating the average number of Hounsfield units per pixel as a percent of the opposite tibia.7,11,37 In-line strain gauges applied to the ring external fixator, during in vivo distraction and subsequently during differential soft tissue release, correlated biomechanical properties to the radiographic and histologic findings.10 Resistance to distraction was highest in the regenerating bone bridge; the residual load across the tibial distraction gap was significantly higher than loads measured in surrounding muscle, fascia, and skin.70 Measured loads increased weekly during distraction, related to rapid mineralization of the collagenous bridge; as distraction increased the gap between bone ends, expanding radiodense projections decreased the radiolucent central zone measured on serial radiographs.6,35,67,69 Distraction loads were proportional to the cross-sectional area of the distraction bridge which correlated to the local host bone anatomy (metaphyseal distraction sites higher than diaphyseal).5,6,7,10,17 Load measurements as early as the second week of distraction could predict premature consolidation of the gap (high load) or future nonunion (low load) attributable to disruption of the biologic bridge. These combined findings resulted in U.S. patent Number 5,437,668 for an in vivo strain gauge system to provide clinical guidelines for distraction.

From these studies emerged a model for the bone-forming unit as the structural basis of distraction osteogenesis (Fig 1A). Under standardized conditions, the distraction gap appears to be bridged by numerous microscopic bone-forming units (parallel to the distraction force), each undergoing similar cellular and matrix transformations correlating to radiographic, histologic, angiographic, and biomechanical findings (Fig 1B). In this bone-forming unit, each longitudinal microcolumn of new bone is directly accompanied by the growth of adjacent sinusoidal vessels.2,6,25,34 Trueta64 originally described the intimate association between blood vessels and bone formation. The juxtaposition of osteoblasts and vessel lining cells may be critical to direct bone formation.23

Fig 1.

Fig 1.

Assuming that osteogenesis and angiogenesis are coupled, how are the progenitor cells attracted, organized, and induced to proliferate and to differentiate in the distraction gap? It is likely that these cells are modulated by local biomediators and may communicate with each other by expression of similar proteins. The bone-forming unit in distraction osteogenesis is proposed to test the hypothesis that fibroblast growth factor-2 (FGF-2) is involved in the molecular control of direct bone formation, because FGF-2 is a potent mediator of cellular chemotaxis, proliferation, and differentiation toward osteogenic and angiogenic cell lines.24,41,59,61

To test hypotheses related to the molecular control of direct bone formation, the bone-forming unit had to be reproduced in a model that would allow molecular characterization and modulation. In situ hybridization to show active expression of mRNA and immunohistochemistry to show protein synthesis have been used in rodent bones. Delivery systems for various biomediators including dietary, hormonal, and protein have been established in rats. Could the distraction osteogenesis model be scaled down and standardized in the rat for testing these hypotheses? Are there any conditions in the rat that might simulate poor bone formation where FGF-2 could be tested as a critical biomediator?

This research followed six phases: reproducing standardized radiographic, histologic, and biomechanical features of distraction osteogenesis in the rat model of tibial lengthening; application of immunohistochemistry, in situ hybridization, and reverse transcriptase-polymerase chain reaction (RT-PCR) to the distraction gap in the rat model to characterize and measure cellular functions during direct bone formation; development of different delivery systems in the rat for modulation of bone formation; use of positive and negative modulators of bone formation in the rat model; evaluating older rats for potential deficits in bone formation using this model; and evaluating the potential role of FGF-25 in mediation of bone formation in young and old rats.

Back to Top | Article Outline

MATERIALS AND METHODS

One thousand eighty-nine rats were used in a series of 70 separate experiments. Power studies for the different experiments required different numbers of rats per group, depending on which parameter was being compared: histomorphometry, radiodensity, biomechanical testing, immunohistochemistry, in situ hybridization, or RT-PCR. Except for the initial pilot studies and mRNA analysis, rats were tested in groups of six to 10.

The body weight was monitored closely in rats to confirm feeding and expected growth. The rats were handled for at least 1 week before surgery and housed in individual cages at 22°C and 50% humidity in controlled rooms having 12 hours of light and 12 hours of dark cycles with lights on at 0600 hours. Surgery was done using semisterile conditions with the animal under anesthesia induced by an intraperitoneal injection of pentobarbital (50 mg/kg body weight). The surgical site was shaved and prepared with 70% ethanol solution. Sterile drapes, gloves, and instruments were used. The electric drill was disinfected with alcohol.

The rats were monitored regularly by independent animal care providers and veterinarians. Postoperative recovery was facilitated in a heated cage with analgesics (butorphanol, 0.1 mg/kg body weight) and antibiotics (topical furazolidone for the tibia or garamycin, 0.25 mg intramuscularly after intragastric intubation). Twice daily inspections for manual distractions, monitoring water and food intake, and observation of walking were recorded. With the animal under anesthesia, sacrifice was done by lethal injection or by decapitation if trunk blood was required.

Primarily Sprague-Dawley rats were used. These virus-free rats were obtained at requested ages and weight ranges. Males were used for all of these experiments to avoid variables associated with the estrus cycle of female rats, except for two experiments testing the effects of ovariectomy. As in the canine model, one tibia was used for surgery with the opposite tibia being the control. In later experiments old and young rats were compared. For very old rats (24 months), CAMM Sprague-Dawley retired breeders initially were used. In a subsequent series of experiments to validate the aging effect, Harlan Sprague-Dawley rats at 12 months versus 4 months of age were used.

Radiographs were obtained at sacrifice using a high resolution closed system at standardized settings (35 kV for 20 seconds).63 Surgically treated and contralateral control tibias were harvested and dissected of extraneous soft tissues to preserve the entire bone and gap healing tissues for immediate, fresh, paired radiographs, carefully orienting the bones in a standard fashion. The radiographs later were digitized using a microscope with a standardized light source and a video camera connected to a CPU with video-grabbing software for real-time focus adjustments. The digitized images then were incorporated into NIH Image Analysis 1.49 (NIH, Bethesda, MD) for various measurements including relative or percent gap density (gray scale, 0–256) controlled by adjacent host bone, opposite side, and percent gap area of new bone using the adjacent medullary canal as a threshold density (Fig 2). After measurements were recorded, the data were divided into experimental groups and statistically analyzed using Student’s t test and ANOVA.

Fig 2.

Fig 2.

For standard hematoxylin and eosin histologic staining, each specimen was fixed in neutral buffered formalin, decalcified in 5% formic acid, cut to include portions of proximal and distal host bone equivalent to the distraction gap, marked with india ink on the proximal side, paraffin embedded, and sectioned to 5–7 μm. Serial sections in the coronal plane were made until a midcoronal plane was reached as evidenced microscopically by the presence of both cortices and the medullary canal of proximal and distal host bone encompassing the bridging distraction gap. This standardized method allowed selective measurement of periosteal or endosteal bone production (separated by the four cortical margins) and selective comparison of proximal versus distal halves of the distraction gap (separated by the central fibrous interzone). The microscopic images were oriented in standardized position, digitized in similar fashion to the radiographs, and measured for the percent of new bone area relative to the total distraction gap area. The distraction gap was measured by a region of interest (ROI) incorporating the four external cortical corners and measured in square millimeters after calibration of the microscope with a micrometer. The area of new bone was outlined by free-hand ROI based on recognition of bone matrix using cellular morphologic features, stain color, and microstructure; the new woven bone could easily be distinguished from fibrous tissue, hematoma, muscle, comminuted microfragments of lamellar bone, fat, or small islands of cartilage. The area of new bone formation was quantitated in square millimeters as endosteal new bone or periosteal new bone (proximal or distal). Endosteal new bone was located between the cortices and bridging the gap ROI. Periosteal new bone was located beyond the cortices and also bridging the gap.

For immunohistochemistry or in situ hybridization, cell counting techniques were used. A standardized microscopic grid was applied to each specimen and relative numbers of stained cells were counted manually.18 The grids were localized to previously described zones of bone formation: fibrous interzone, primary matrix front, proximal or distal, and microcolumn formation, proximal or distal. These zones were localized in the gap to either endosteal new bone or periosteal new bone.

Mechanical testing of the distraction gap was done in vitro.10 In vivo strain analysis as published in dogs10 was not attempted in the rats. However, a special method for testing the strength of the nascent bone bridge during the early transition from active distraction to consolidation was developed. The ultimate tensile strength, tensile stiffness, and energy to failure were measured.

Immunohistochemical analysis was done using proliferating cell nuclear antigen (PCNA) to determine prevalence of cell division. This method was compared with H3 thymidine labeling as a gold standard with high correlation before being accepted for study.18

Various biomediators also were immunostained with appropriate antibodies including the primary focus of this work, FGF-2 and its receptors (FGFR-1, 2, 3). Sections were deparaffinized in xylene, rehydrated in ethanol, washed in phosphate buffered saline (PBS)/0.02% Triton, exposed to Citra (Antigen retrieval Citra, Biogenex, San Ramon, CA) at 95°C for 15 minutes to retrieve the epitope, incubated with 2% bovine serum albumin at room temperature to block unspecific binding, and exposed to levamisole for 30 minutes to block endogenous alkaline phosphatase (ALP) activity. The sections then were exposed at 4°C overnight to one of the following affinity-purified rabbit polyclonal antibodies (Santa Cruz Biotechnology, Santa Cruz, CA) diluted 1/50: antibody against a peptide corresponding to amino acids 808–822 mapping at the COOH terminus of the human Flg gene product for detection of FGFR-1; antibody against resides 805–821 mapping in the carboxy terminus of the human bek receptor for detection of FGFR-2; or antibody against a peptide mapping in the amino terminal domain of FGF-2 of human for detection of FGF-2. Control sections were incubated with rabbit immunogamma globulin (Vector Laboratories, Burlingame, CA) at the same concentration as that of primary antibody. All sections were washed and exposed to a second antirabbit biotinylated antibody (1/100) at room temperature for 30 minutes. After washing with PBS, the sections were exposed to avidin-biotin alkaline phosphatase complex (ABC) at room temperature for 30 minutes, and then were revealed by exposure to chromogen, counterstained with eosin.

In situ hybridization using biotinylated oligomers was done using an mRNA detection system (Biogenex, DA100-SS). The unstained sections were deparaffinized through xylene to distilled water and allowed to dry completely at room temperature. Deproteinization was done using proteinase K (0.2 mg/mL) at room temperature for 15 minutes. The slides then were washed, dehydrated, and dried at room temperature for 15 minutes. Biotinylated probes were diluted to 250 ng/mL in hybridization solution containing formamide and sodium chloride/sodium citrate buffer and placed on the tissue sections. The slides were heated to 95°C for 8 minutes to reduce mRNA hairpin loops and folding. The slides were placed in a humidity chamber for hybridization overnight at room temperature. The slides were washed, incubated with normal goat serum (20 minutes), washed, incubated with mouse antibiotin antibody (20 minutes), washed, incubated with goat biotinylated antimouse antibody (20 minutes), washed, incubated with ALP-conjugated streptavidin (20 minutes), washed, incubated with nitro blue phosphate/toluidine salt substrate (20 minutes or until color developed), washed, dehydrated to xylene, dipped in toluene, and mounted in protex. Multibiotinylated52 oligomeric probes were rat FGF-2, poly dT mRNA positive control, and mutated poly T negative control (Invitrogen, Carlsbad, CA). Also included was one 5′ biotinylated poly dT mRNA control (Biosynthesis, Lewisville, TX). The initial experiments compared Brigati-labeled 3′ multibiotin (Invitrogen) versus single 5′ biotinylated (Biosynthesis) poly dT oligomers to set up the initial conditions and to test for probe penetration into all zones of the distraction gap.26,52 Good specific binding was obtained to Brigati and single biotinylated probes. The Brigati-tagged probes were more sensitive as measured by the time required for the substrate to develop color. The pattern of poly dT in situ hybridization indicated equal penetrance of probe into all zones of the distraction gap. To test for specific mRNA transcripts, a rat FGF-2 antisense oligomer with a mutated poly T was used as a negative control (Brigati-tailed probes). The antisense sequence was complementary to the sense sequence from + 752–783 (5′-AAG GGA GTG TGT GCG AAC CGG TAC CTG GCT-3′).65 The mutated poly T control consisted of poly T sequences interrupted every three thymidines by GCGC. It has been shown that antisense FGF-2 RNA transcripts are expressed physiologically and therefore using a sense rat FGF-2 oligomer as a control was not deemed appropriate.61 Specific binding was obtained to the FGF-2 oligomer as judged by the control probes and by comparison with a digoxigenin-tailed FGF-2 probe of serial slides by an independent laboratory.26

Gap mRNA analysis using RT-PCR was done.48 Total RNA was isolated by using Tri-Reagent (Molecular Research Center, Cincinnati, OH) from the distracted callus collected from lengthened rat tibia at the day of sacrifice. The RT-PCR was done according to the instructions of the manufacturer. Briefly, first-strand cDNA was synthesized by using the THERMOSCRIPT™ RT-PCR System (Invitrogen) with extracted total RNA. The primers specific for rat glyceraldehyde 3-phosphate dehydrogenase (GAPDH) gene were used to equalize the amount of cDNA in the each sample. The RT-PCR products are run on a 1% agarose gel and identified. Every RT-PCR was repeated at least twice with different samples.

Different methods to deliver agents to the rats were tested. The laboratory at my institution has extensive experience with preparing rats for intravenous or intragastric delivery of nutrients.20,58,65 Early experiments with distraction osteogenesis in rats actually tested different nutritional situations with controlled diets. These skull-based catheter systems also were used to administer various toxins such as ethanol. Human recombinant growth hormone was tested through intravenous pulses. Alzet miniosmotic pumps (Durect Corp, Cupertino, CA) were implanted subcutaneously on the back for systemic or in the ipsilateral thigh with a catheter tunneled to the distraction site for localized delivery of biomediators. The 7- or 14-day osmotic pumps were used for continuous delivery of FGF-2 (Scios, Fremont, CA). Local delivery from the osmotic pumps could be directed to either extraperiosteal or intramedullary locations in the tibia.

Delivery by direct intramedullary injection also was done. Microangiographic techniques were used in vivo to follow the path and distribution of injected iodinated contrast agent with high resolution radiography and post vivo injected barium sulfate dye for histologic analysis. Using an anterior unicortical drill hole angled distally and a special syringe, microvolumes of commercially prepared FGF-2 in a hyaluronic acid carrier45 (Ossigel, Orquest, Mountain View, CA) were injected at the time of surgery, before the low energy osteotomy.

To confirm delivery of exogenous agents, enzyme-linked immunosorbent assay (ELISA) was used on the rat’s blood collected at sacrifice. Trunk blood was allowed to clot for 30 minutes before centrifugation for 30 minutes at approximately 2560 times gravity. The serum samples were stored at −2°C and thawed once for assay. The samples were assayed for FGF-2 protein with a Quantikine HSFBO microtiter plate assay (R&D systems, Minneapolis, MN) after the bench-top procedure. Briefly, 150 μL of standard or sample was added to each well coated with monoclonal antibodies to recombinant human FGF-2 and incubated for 20 hours at room temperature. After extensive washing, an ALP-conjugated polyclonal antibody to FGF-2 was added. After washing to remove any unbound antibody-enzyme reagent, a substrate solution (NADPH) was added. After a 60-minute incubation an amplifier was added for 30 minutes and color developed in proportion to the amount of FGF-2 bound in the initial step. The optical density of each well was determined within 30 minutes using a microtiter plate reader set to 490 nm for the readings and to 650 nm to correct for imperfections in the plate. This immunoassay recognizes natural rat, human, and escherichia coli-derived recombinant FGF-2. The assay sensitivity is 0.28 pg/mL with a range of 0.5–32 pg/mL for serum samples.

Back to Top | Article Outline

RESULTS

In Phase 1, the hypothesis that a scaled rat tibial lengthening model would replicate the radiology, histology, and biomechanics of distraction osteogenesis found in the canine tibial lengthening model was tested.

Five separate experiments using 103 rats were done to compare radiographic, histologic, and biomechanical findings with those in previous work using the canine model. Variables tested included the size, material, and configuration of holes in rings, size and type of transosseous wires, wire fixation bolts or screws, method of osteotomy, distraction rate, latency period, and temporal sequence. The goal was to replicate the temporal and spatial zones of the histologic features seen in dogs.

Rat handling required housing in individual plastic cages to avoid problems among rats or between the eternal fixators and the wire cages. Intraperitoneal injection of sodium pentobarbital (50 mg/kg) was found to be reliable with more than 95% survivorship in all experiments.

After many evolutions in the hardware and technique, the surgical procedure eventually became standardized and highly successful. The first step involved manual local thumb compression to fracture the fibula. Next the steel ring fixator was sized for the tibia by adjusting the length of the two distraction units to position the proximal ring below the knee and the distal ring above the ankle. The fixator then was used as a template to drill crossed transosseous wires (22-gauge, 3.5-inch spinal needles) at 45° (one above and one below each ring). The predetermined wire sites were chosen carefully to avoid local vessels and nerves. Each wire was tensioned to the ring by twisting the two connecting bolts.1,12,13 The low-energy midtibial osteotomy was produced by percutaneously drilling a 1.0-mm K wire across the midtibia and then manual three-point bending to fracture the tibia. When open techniques were required for placement of a delivery catheter, wound closure was done with silk sutures.

Wound problems were rare. Minor weight loss was observed during distraction, despite eating and drinking. Pain did not seem to be an issue. In early trials, there were a few isolated cases of contracture of the ipsilateral toes, indicating inaccurate wire placement across muscles or a nerve. The majority of rats bore weight on the surgically treated leg with a slight limp. The fixator did not cause skin breakdown and remained free of urine and feces. The rats did not claw or bite the affected leg.

The rat model19 is similar to the canine model in that a two-ring external fixator with two tensioned transosseous wires per ring and a low-energy midtibial osteotomy were used. The fixator size was scaled down to the rat tibia (40 mm), approximately 20% as long as the canine tibia (200 mm) (Fig 3). The optimal rate of distraction eventually was determined to be 0.4 mm/day (range, 0.2–1 mm) at a rhythm of 0.2 mm twice daily and the optimal latency of 1 day (range, 0–7 days), which maximized intramembranous ossification in the gap, comparable to the canine model. The distraction rate was scaled down 40% from 1 mm/day (dog) to 0.4 mm/day (rat) maintaining a rhythm of two increments per day. The latency period was reduced to 1 day (rat) from 7 days (dog). The absolute linear distraction gap was decreased from 28 mm in the dog to 7 mm in the rat while retaining 15% lengthening. This 7-mm distraction allowed for the entire gap to be observed at the ×1 objective, microscopically.

Fig 3.

Fig 3.

Using a nondistracted osteotomy (fracture model) for comparison, the radiographic and histologic findings were tracked and compared with time for the distraction period (14–20 days) and the consolidation period (70 days). Rats (like rabbits) tend to react to a fracture with voluminous cartilage formation that later undergoes endochondral ossification. Using percutaneous drilling and manual osteoclasis compared with open sawing, a shorter latency, and a distraction rate of 0.2 mm twice daily, cartilage formation was minimized and direct bone formation was maximized in the distraction gap.32 The central fibrous interzone with mirror image transitional zones proximally and distally to microcolumns of new bone and interspersed vascular channels was achieved in most young rats (4–8 months). Progressive ossification, bridging (between Days 20 and 30), consolidation (by Day 50), and remodeling (by Day 70) of the distraction gap occurred reliably (Fig 4).

Fig 4.

Fig 4.

A method for in vitro mechanical testing of the distracted tibias was developed in a series of experiments involving 66 rats. Although bone usually is tested in torsion or in four-point bending, these methods could not be applied reliably to early bone formation at the soft collagen stage.68 Tendons and ligaments (pure collagen) routinely are tested in tension. Therefore tensile testing was compared with bending tests and it was found that tensile testing could significantly distinguish initial bridging of the distraction gap and show the progressive increase in strength associated with radiographic remodeling. The results tracing the development of tensile strength from the end of distraction (Day 20) to complete consolidation and remodeling (Day 70) have been reported.15 The ultimate tensile strength rose linearly from Days 20–70 with the most significant increase between Day 20 and Day 30, at radiographic bridging (Fig 4). Tensile stiffness increased significantly only at bridging but did not change during remodeling. Energy to failure increased progressively from Day 20 to Day 70. Ultimate tensile strength at Day 70 revealed that the tibias were more than 90% as strong as the normal, contralateral controls. However in three-point bending at Day 70, the distracted tibias were only 65% as strong as controls, similar to the canine experiments.10,15

Pulsed fluorescent probes (tetracycline labeling), were used to follow the pattern of mineralization in the osteoid of the microcolumns of new bone. Fluorescence microscopy showed that the microcolumns mineralized from deep to superficial and from host bone surfaces toward the tips of each new microcolumn. Nondecalcified specimens were cut, ground to 50 μm, and stained with von Kossa silver stain to confirm the tetracycline patterns. Selected specimens following tensile testing to failure were examined by fluorescence microscopy revealing that failure occurred at the central fibrous interzone.

In Phase 2 it was hypothesized that immunohistochemistry, in situ hybridization, and RT-PCR could be used to characterize and semiquantitate cellular activity at different stages of distraction osteogenesis.

The findings of standard histology in this model (Fig 5) are important to describe to relate findings of immunohistochemistry and in situ hybridization. The central region of the distraction gap is radiolucent on radiographs during active distraction. Histologically, this zone corresponds to parallel bundles of collagen, in line with the distraction force, called the fibrous interzone. On both sides of the fibrous interzone, important junctional zones connect the unmineralized collagen to the nascent bone columns on either side.66 These two mirror image proximal and distal junctional zones include concentrations of cell proliferation (shown by tritiated thymidine and PCNA immunohistochemical analysis) located at the site of initial bone matrix. They are termed primary matrix fronts. The tips of each new bone column expand to maximum diameters of 150 to 200 μm, typical of trabecular bone. As they extend toward each host bone surface, each microcolumn is surrounded by thin-walled sinusoidal vessels of equal or slightly larger diameter, all parallel to the distraction force. These mirror image proximal and distal zones are termed the zones of microcolumn formation.

Fig 5.

Fig 5.

Microscopically, the cell types in the fibrous interzone and microcolumn formation zones are morphologically distinct. Osteoblasts, fibroblasts, and endothelial cells are easily distinguished by their shape, size, configurations, alignment, and associated matrices.

In contrast, the zones of the primary matrix front are filled with proliferating cells and early precursors that have been more difficult to differentiate by routine histologic analysis. Immunohistochemical stains were used to identify various proteins in the cytoplasm or in the matrix adjacent to these cells. In situ hybridization was used to identify the intracellular mRNA, which has the potential to produce these proteins. Gap mRNA analysis using RT-PCR was used to semiquantitate the specific genetic constituents during distraction. Using stains for specific proteins and protein receptors on contiguous histologic sections provided insight into cell-to-cell communications.

Cell-counting techniques, manual and automated, were developed to analyze positively stained cells in a grid of fixed ROI (Fig 6). In this way, cellular activity could be evaluated within zones: fibrous interzone, primary matrix fronts, microcolumn formations, and marrow compartments proximal and distal.

Fig 6.

Fig 6.

Proliferation patterns in the distraction gap were compared with patterns in a nondistracted fracture.18 The gold standard, uptake of the radioisotope H3 thymidine, was correlated to PCNA immunohistochemistry to standardize the cell counting methods. The numbers of proliferating precursor cells in the adjacent proximal and distal endosteal marrow spaces peaked at Day 2 in distraction and fracture groups. Correlation between the methods was high.

Although bone and cartilage matrix usually could be differentiated by cell morphology and uptake of hematoxylin or eosin stains, Collagen I and II stains were used to confirm the matrix and cell types.66 Additional confirmation of osteoblast cell types was accomplished using osteocalcin stain. Osteogenic precursors could not be distinguished reliably using Cbfa-1 immunohistochemistry.

Various growth factors associated with the transition of collagen to bone were used to establish the specific activity of osteoblasts at different stages of intramembranous ossification. Transforming growth factors (TGF-β1, TGF-β2, and TGF-β3) and bone morphogenetic proteins (BMPs) initially were selected for in situ hybridization.48,50 Although TGFb in situ hybridization was positive at the transition zone from fibroblasts to osteoblasts and BMP-2 in situ hybridization was positive in osteoblasts in the new bone, the mRNA for FGF-2 was found to be specifically upregulated in osteoblasts and vessel lining cells along the new microcolumns of bone (Fig 7). Additionally, FGF-2 and three of its high affinity receptors (FGFR-1, -2, and -3) were identified using immunohistochemistry, extending into the new bone columns and the new blood vessels in the zones of microcolumn formation (Fig 8). Based on the immunohistochemistry and in situ hybridization of hundreds of specimens from rats having distraction osteogenesis, FGF-2 was the most strongly correlated growth factor to the early transition to osteogenic and angiogenic cell types.

Fig 7.

Fig 7.

Fig 8.

Fig 8.

The findings from immunohistochemistry and in situ hybridization were supported using semiquantitative RT-PCR. Strong bands of mRNA for FGF-2 were found in tissue taken from the distraction gaps. The housekeeping gene GAPDH was used as a control to standardize the relative intensity of each band (Fig 9).

Fig 9.

Fig 9.

In Phase 3, it was hypothesized that different delivery systems could be used in the rat model of distraction osteogenesis to deliver bioactive substances either systemically or locally.

These interventional techniques were developed to modulate the local osteogenesis, positively and negatively. Using systemic or local delivery systems, bioactive substances could modulate the local processes, measured in the distraction gap. Problems to overcome included targeted delivery, dosage, reliability, clinical feasibility, side effects, and tolerance. In seven experiments, 160 rats were used to test these delivery systems.

Dietary factors were examined initially, because the rats routinely lost weight (10%) during the distraction phase. The tibial lengthening model was combined with a previously standardized total enteral nutrition model in the same rat.17 Under the first anesthetic, the rat had separate cannulae placed surgically into the femoral vein and into the stomach. The cannulae were tunneled subcutaneously to the dorsal neck where they were attached to a special headpiece. Using a stereotactic skull holding device, the posterior occipital bone was exposed, five jeweler’s screws were placed to anchor acrylic dental cement, and the cannulae were secured. A swiveling spring-suspended tethering device connected the cannulae to supply lines at the top of the cage (Fig 10A). Each rat was able to walk freely around the cage while receiving an intragastric diet. The venous catheters remained patent for at least 3 weeks, allowing blood samples to be withdrawn. The rats received isocaloric diets for 7 to 10 days of acclimatization before the second surgery for placement of the tibial external fixator for lengthening. The combined models were well tolerated and remained functional for the duration of the experiments without apparent side effects or toxicity.

Fig 10.

Fig 10.

Different sized Alzet miniosmotic pumps (15–20 mm long × 5–7 mm diameter, enclosed in Silastic capsules) are commercially available for continuous delivery of soluble agents for either 7 or 14 days. Initially the pump was implanted dorsally near the neck in subcutaneous tissues for systemic delivery. In later experiments the pump was implanted subcutaneously in the lateral thigh, with an attached Silastic catheter tube. This delivery tube was tunneled subcutaneously to the midtibia (Fig 10B), where it was sewn to the anterior fascia adjacent to the distraction osteogenesis site, for extraperiosteal delivery to the distraction site. These pumps were implanted with the animals under the same anesthetic as was used during placement of the external fixator and osteotomy. These pumps require priming for several hours before implantation to deliver the agent immediately and continuously. The pumps were well tolerated without inflammatory reaction, infection, or pain. They successfully delivered greater than 95% of the vehicle with only a small fraction remaining in the pump when disassembled at sacrifice.

Standard techniques for ELISA were used to detect and quantitate serum levels of FGF-2. Radioimmunoassays (RIA) on serum were used to detect insulinlike growth factor (IGF- 1).17 These assays were done routinely in triplicate with quality controls.

The technique for direct intraosseous delivery was developed. At the time of surgery, the external fixator was placed and then the site of the osteotomy was predrilled with a unicortical hole in the anterior cortex, angled distally. A Hamilton microinjection syringe (Hamilton Company, Reno, NV) was positioned in the hole for select delivery of microvolumes. Microangiograms were done in vivo with high resolution imaging equipment. Distribution and local retention were tracked. Trial injections with radiopaque materials of different viscosity (barium sulfate) were done to determine in vitro distribution. Local retention was improved by increasing the osmolarity and viscosity, although some venous run-off was encountered. Volumes exceeding 50 μL extruded from the bone into the surgical site. Using microangiography and barium sulfate injection histology, the intramedullary space was measured to be 25–50 μL in the rat tibia. Histologically, no inflammatory reaction was seen in the medullary canal although the distal marrow contents occasionally were disrupted.

In Phase 4, it was hypothesized that positive or negative modulators of bone formation could be delivered to enhance or diminish bone formation by distraction osteogenesis in this rat model.

Toxins, simulation of disease states, or inhibitors of normal osteogenic processes were introduced to better understand which factors are critical to the local biologic features of intramembranous ossification. Positive modulators of osteogenesis such as diet, hormones, or biomediators were administered to potentially overcome deficiencies found in disease or to accelerate the normal physiologic processes. In 16 experiments, 480 rats were used to deliver systemic modulators or simulate a deficiency in systemic regulators to measure the effects on distraction osteogenesis.

Using the gastric cannula system, groups of rats having tibial lengthening were compared for body weight, bone formation, and serum IGF-1. The total enteral nutrition diet included L-amino acids, minerals, and vitamins infused at rate of 160 kcal/kg/day meeting nutritional requirements (protein 16%, carbohydrate 74%, and lipids 10%) of the National Research Council.22,49 The results determined that total enteral nutrition stimulated weight gain during distraction (compared with weight loss on a chow diet) and resulted in significantly better bone formation by radiodensity and radiographic and histologic bone area without any change in endogenous levels of IGF-1. Neither the positive effect of weight gain nor the negative effect of weight loss on bone formation seemed to be mediated by circulating levels of IGF-1.49

In another experiment, the effects of diminished (from the normal 1% down to 0.125%) concentrations of calcium and phosphorus in the total enteral nutrition diet were measured. No significant differences in bone formation were found. Presumably, the rat skeleton contained adequate stores of calcium and mechanisms for mobilizing this calcium to provide adequate blood levels of these minerals for incorporation into the newly formed bone bridge. Acute deficiency of dietary calcium and phosphorus did not inhibit the distraction osteogenesis process in young, growing rats, otherwise receiving a normal total enteral nutrition diet.

In a small pilot study, ovariectomy was done in female rats to evaluate estrogen deficiency (osteoporosis Type 1) on distraction osteogenesis. The combined surgeries were successful, but no acute decrease in bone formation was seen during distraction osteogenesis in these relatively young female rats. Either ovariectomy requires more time to affect bone metabolism or the younger female rat can overcome the hormonal deficiencies of acute ovariectomy by other compensatory mechanisms.

The consequences of switching the protein source for rats reared on a defined AIN-93G diet were examined. Either casein as the sole protein source or two other semipurifed AIN-93G diets in which the protein source was switched to either soy protein isolate or rice protein isolate were tested. The latter two diets contain phytochemical factors associated with the protein fraction. Female rats had their diets switched at age 60 days and were allowed to consume a new diet ad libitum for 3 weeks. Half of the groups also were ovariectomized. Newly acquired imaging equipment for peripheral quantitative computed tomography (pQCT) for rat bone was used to analyze trabecular bone in the proximal tibia. Short-term ovariectomy resulted in a significant loss of trabecular bone in casein-fed controls and rice protein isolated rats (p < 0.05). Neither rice nor soy protein isolate diets had significant effects on bone mineral density in intact animals. Only soy protein isolate prevented the ovariectomy-associated loss of trabecular bone (p < 0.05). This suggests that soy protein isolate contains factors which support bone growth in the absence of endogenous estrogens. It is likely that these factors include the soy-associated phytoestrogen isoflavones genistein and daidzein.

Different dietary toxins have been tested. Lead exposure historically has been related to adverse health effects in young children from low birth weight to decreased growth rates and reduced stature. Lead ingestion in children can be directly correlated to time of ingestion by lead deposits within growth lines on standard radiographs. Impregnated Sprague-Dawley rats, from gestation Day 4, were exposed to lead acetate in differing concentrations (0, 825, or 2475 ppm) with acetate controls in their drinking water. Elevated blood levels of lead were confirmed in the litters of male and female Sprague-Dawley rat pups. Intact tibias were tested by three-point bending strength with a dose-dependent decrease in load to failure, only in male pups. No differences in plasma levels of vitamin D metabolites could be measured. When lead exposure was extended into puberty, dose-dependent differences in somatic and long bone growth were seen. In a second lead experiment, pubertal pups exposed to lead were given saline, L-dopa, testosterone (males only), dihydrotestosterone (males only), or estradiol (females only) by Alzet pumps. Sex steroid replacement did not restore skeletal strength in lead exposed pups. L-dopa increased the plasma IGF-1, bone growth, and bone strength in control rats but had no effect on the lead-exposed pups. In postpubertal, 100-day-old rats exposed to lead, tibial distraction osteogenesis was done. Despite being relatively young, the total new bone was significantly reduced by lead exposure. Lead has been shown to inhibit vitamin D3 stimulated synthesis of osteocalcin by osteoblasts.57

Chronic alcohol abuse has been shown to decrease bone mass, inhibit osteoblastogenesis, increase fracture risk, and delay fracture healing.20,22,29,46,47 Seventy-nine 3-month-old male Sprague-Dawley rats were tested for alcohol effects in sequential studies. Three-point bending strength was reduced significantly in the tibias of rats exposed to chronic ethanol intragastrically. The tibias were structurally weakened with significantly lower trabecular volumetric bone mineral density and cortical cross-sectional area by pQCT. When distraction osteogenesis was combined with chronic ethanol ingestion, bone formation was reduced significantly. The ethanol or its metabolites may have a direct effect on the process of bone formation, because all rats were given total enteral nutrition to offset any direct nutritional effects.

Using the preimplanted intravenous catheter system and the distraction osteogenesis model, exogenous recombinant human growth hormone (rhGH)21 was administered without measured acceleration of bone formation in young rats. Using Alzet pumps, exogenous dihydrotestosterone (dHT) was administered to males and estrogen to females having distraction osteogenesis. No positive effect could be measured in these young rats. It is likely that the process of bone formation during distraction osteogenesis in young rats already is at maximum rates of cellular proliferation and differentiation, which cannot be accelerated additionally by exogenous hormonal supplementation.

In Phase 5, it was hypothesized that distraction osteogenesis would be reduced measurably in older rats relative to younger controls.

The effect of age on bone formation during distraction osteogenesis was studied by comparing rats (n = 59) grouped by age. The results of the first 30 rats were published.8 To maximize the potential differences between young and old rats, very old (24 months) retired Sprague-Dawley breeders versus young rats (9 months) from CAMM were used. A significant reduction (p < 0.003) in mineralized bone by radiographic comparison (51% versus 68%) was seen. The experiment was repeated with Harlan Sprague-Dawley rats, comparing rats 24 months (old) with rats at 3 months (young) and an even more significant (p < 0.001) difference was seen in radiodensity (34% versus 95%) in the distraction gap. Histologic analysis was done on the Harlan rats. Periosteal new bone formation was similar in old and young rats, but endosteal new bone formation was reduced in old rats to ¼ that in young rats (p < 0.01). In old rats, the periosteal new bone formed by normal zones of intramembranous ossification, whereas the endosteal gap was filled with primarily a loose fibrous matrix similar to the fibrous interzone (Fig 5B). This loose fibrous matrix had spindle cells (fibroblasts) and collagen but no new trabeculae of bone. Sporadic vessels were seen. The junctional zone of concentrated cells leading to coupled osteogenesis and angiogenesis was not apparent. Proliferating cell nuclear antigen stains revealed concentrations of proliferating cells peripherally at the periosteal new bone. Positive PCNA cells were present in the endosteal distraction gap but were widely scattered, never forming a concentrated front. Mesenchymal cells retained proliferative potential but could not organize or differentiate. Failure of proliferating cells to organize and differentiate might be related to a deficit of FGF-2.

Other features of the rats’ tibias were examined during aging. The tibias grew in length and diameter but did not show a significant increase in cortical thickness. The contents of the bone marrow cavities were examined proximally and distally using density thresholds to measure relative nuclear area and fat area. Younger rats had a significantly higher ratio of nuclei to fat cells compared with older rats. Aging is associated with increasing fat replacement of nucleated cells in the marrow.43

In Phase 6, it was hypothesized that the transformation from fibrous tissue to microcolumns of new bone seen during distraction osteogenesis in young rats is strongly correlated to the local expression of FGF-2. Conversely, the failure of bone formation from fibrous matrix seen during distraction osteogenesis in old rats is related to a local deficiency of FGF-2.

The dramatic deficiency in old rats having distraction osteogenesis (Phase 5) of an organized zone of coupled angiogenesis and osteogenesis leading to bone formation led to exploration of the role of FGF-2 and its receptors. Because FGF-2 is recognized as one of the most potent stimulators of proliferation, chemotactic organization, and differentiation into osteoblastic and angioblastic cell lines, it was hypothesized that the endosteal new bone formation in young rats would show intense staining of FGF-2 and its receptors at the transitional zones where coupled osteogenesis and angiogenesis occur. Additionally it was thought that FGF-2 or its receptors would be reduced dramatically in older rats. To date, 196 rats have been used in this phase of the study.

In young rats, the expression of FGF-2 and its high affinity receptors FGFR-1,-2, and -3, were shown by immunohistochemistry, in situ hybridization, and RT-PCR. As a control, the adjacent proximal tibial growth plate showed FGF-2 in resting zone cells and FGFR-3 deep in the cartilage cell columns, whereas FGF-2, FGFR-1, and FGFR-2 were present at the primary and secondary bone trabeculae, in the vicinity of sinusoidal vessels, osteoblast concentrations, and new bone formation. In young rats having tibial lengthening by distraction osteogenesis, FGF-2 was intensely expressed in cells in the zone of proliferation, extending along new vessels and along new bone microcolumns (Fig 8). The positively stained cells seemed to be primarily precursors and early osteoblasts, as found by other researchers.71 The distribution of FGFR-1 overlapped the FGF-2 cells and extended into osteoblasts in and on the surface of the new bone microcolumns. The distribution of FGFR-2 was more strongly expressed in and around new vessels. Expression of these FGF-2 and its receptors was sparse in the fibrous interzone.

In old rats, the growth plates were closing with reduced staining for FGF-2 and its receptors. In the distraction gap of old rats, the expression of FGF-2 and FGFR-1 were present but decreased, while FGFR-2 and FGFR-3 were present in periosteal bone forming areas. None of these markers was expressed in the loose fibrous bridge in the endosteal new bone areas.

The distraction gaps in young and old rats were harvested for mRNA analysis by RT-PCR. Using the housekeeping gene GAPDH to normalize for other genes, relative levels of FGF-2, FGFR-1, FGFR-2, Cbfa-1 (early osteoblast marker), osteocalcin (later osteoblast marker), and Pecam (early endothelial cell marker) expression were compared in specimens from young and old rats. Each distraction gap specimen included endosteal new bone, periosteal new bone, and all intermediate zones. The semiquantitative analysis showed that only osteocalcin and FGF-2 were reduced in older rats.

In Phase 7, it was hypothesized that if the deficiency of FGF-2 shown in old rats is critical for bone formation, then administration of exogenous rhFGF-2 to old rats should restore endosteal new bone formation to levels seen in young rats.

The effect of systemic administration of rhFGF-2 on distraction osteogenesis in 18-month-old rats was studied in two groups, vehicle (20 mmol/L sodium citrate, 1 mmol/L EDTA, 9% sucrose, pH 5.0) and vehicle + rhFGF-2 4.3 mg/mL solution of trofermin (Fiblast, Scios, Fremont, CA). An Alzet 2002 pump (87 μg/kg/day × 14 days) was primed and implanted between the scapulae, subcutaneously. At sacrifice (end distraction), trunk blood was collected for ELISA analysis for FGF-2. Rats having distraction osteogenesis were used for ELISA controls. The young control rats had a baseline of 6.13 pg/mL, the old control rats had 3.88 pg/mL, and the old rats treated with systemic FGF-2 had 21.73 pg/mL. All mean values were within the expected published ranges for humans, 2–3 pg/mL in adults and 6–8 pg/mL in neonates.72 The vehicle control old rats had diminished bone formation radiographically and histologically, comparable to old rats in previous experiments. In contrast, the old rats treated with systemic rhFGF-2 had improved endosteal new bone formation (p < 0.02), but only at the proximal side of the distraction gap, similar to a previous report.53

Next, the effects of local, extraperiosteally administered rhFGF-2 were tested on distraction osteogenesis in older rats. Using a 7-day Alzet pump (#1007D) placed subcutaneously in the lateral thigh with an attached catheter, tunneled to the distraction site and sutured to the fascia, adjacent to the lateral periosteum (Fig 10B), vehicle or rhFGF-2 was delivered. Four groups were compared: young vehicle, young rhFGF-2, old vehicle, and old rhFGF-2 (Fig 2). The aging effect was shown by comparing the young with old vehicle rats. Histologically, the old vehicle rats had significantly less (p = 0.02) endosteal new bone (15.7%) than young vehicle rats (50.5%). The old rhFGF-2 rats had significantly increased (p = 0.008) endosteal new bone (36.7%) than the old vehicle rats. Histologically, the new bone was improved at proximal and distal endosteal new bone sites, despite a lower percentage of nucleated cells in the distal marrow space (Fig 11).

Fig 11.

Fig 11.

A second mode for local delivery of rhFGF-2 by local intramedullary injection (intraosseous) as described in Phase 3 was tested. The hyaluronic acid carrier for injectable rhFGF-2 was prepared commercially (Ossigel, Orquest).56 The concentration of rhFGF-2 (4.0 mg/mL) from Orquest was similar to that provided by Scios. In a pilot study using 100 μL of Ossigel versus a hyaluronic acid control (Orquest), endosteal new bone during distraction osteogenesis was increased dramatically by Ossigel to bridge approximately 75% of the gap in 18 old rats. In a subsequent experiment testing a range of dosages by varying the injected volumes (25, 50, or 100 μL), 25 to 50 μL were optimal, which matched the calculated volume of the medullary canal. Various other organs (liver, adrenal, kidney, spleen, and lung) were collected and no adverse effects were seen histologically.

The two local delivery systems for rhFGF-2 were compared using distraction osteogenesis in old rats: extraperiosteal rhFGF-2 by Alzet and local intramedullary injection of Ossigel. In two separate experiments, four groups were tested: extraperiosteal rhFGF-2, extraperiosteal vehicle, intramedullary Ossigel, and intramedullary hyaluronate vehicle. The effects were compared radiographically and histologically. Additional specimens were tested mechanically to failure in tension as described in Phase 2. Radiographic analysis of specimens at 14 days (end distraction) revealed that intramedullary Ossigel was greater than intramedullary hyaluronate vehicle (p < 0.001) and extraperiosteal rhFGF-2 by Alzet was greater than extraperiosteal vehicle (p = 0.015). The intramedullary Ossigel was slightly better than extraperiosteal rhFGF-2 by Alzet but not quite significant. The tensile tests were done until failure. Intramedullary Ossigel showed the highest stiffness and energy to failure, which was significant compared with the vehicle. The mean load to failure at Day 7 of distraction (20.4 N) in the intramedullary Ossigel group was nearly twice that in the vehicle (12.3 N) and equivalent to vehicle treated tibias 10 days later at Day 17 (22.4 N).

The improved endosteal new bone in old rats treated with exogenous rhFGF-2 was assessed by immunohistochemistry. The histologic zones of progressive ossification seen in younger rats were reproduced with positive expression of PCNA, FGF-2, and FGFR-1. The effects of rhFGF-2 on local marrow constituents still are being assessed.

Back to Top | Article Outline

DISCUSSION

Rats presented more than problems of size, with actual structural differences, lacking normal Haversian canals in the tibial cortex, even though the individual cell types had similar diameters to those measured histologically in dogs and humans. Growth plates remain active for a much more extended time during the rat’s life cycle than those in dogs or humans, and consequently the rat continues to grow in length and weight for most of its life cycle. The growth and life cycles for the Sprague-Dawley rats used for most of these experiments are documented in the Harlan Product Guide. The body weight usually is a good indicator of the rat’s age. The growth rate slows at approximately 12 months and survivorship falls off sharply at 24 months. Other indicators of maturation include minor structural changes in tibial cortical thickness (decreasing with age) and internal diameter (increasing with age) consistent with observed changes in the human bones with senile osteoporosis.14

Despite the many interspecies differences noted above, the surgical methods were relatively well-tolerated and successful for achieving a 15% tibial lengthening that fully bridged with normal bone. Similar to rabbits, early osteotomy healing in rats, before distraction, was found to contain large amounts of cartilage in the fracture gap, which ossified by endochondral and transchondral mineralization.42 By eliminating the latency period, immediate distraction at the optimal rate rapidly transformed the bone-forming cells to a more consistent intramembranous pattern of mineralization. Numerous investigators have discussed the prevalence and importance of endochondral ossification in models of fracture or osteotomy healing.42 The goal of the current model, however, was to produce primarily intramembranous ossification, comparable to that shown in the dog, to test hypotheses relative to the molecular control of osteogenesis and angiogenesis. From observations in nearly 100 dogs2,4,6,7,9,10,14,17 and currently more than 1000 rats, distinct zones of cellular and matrix transformation into bone by primarily intramembranous bone formation were seen.8,15,18,19 These histologic features were reported by others as well.25,28,44,60,62

The radiographic methods used to measure bone formation were standardized to contralateral, nonsurgically treated tibias based on careful measurements of photodensitometry using aluminum step wedges in the earlier canine experiments. Using the NIH Image Analysis software to determine the relative gray density of cortical and cancellous bone, the gap tissues could be reliably and reproducibly measured. Two methods have been used, percentage area of new bone in the distraction gap and average density per pixel in the distraction gap. The former method correlated better to histologic analysis. These methods are labor intensive but reliable. Automated systems, in particular pQCT for small animals, are now more commonly available and probably will replace the radiographic methods used the current study.

Mechanical testing presented special challenges. The distraction gap tissues were partially rigid (mineralized bone) and. partially soft (unmineralized collagen, like tendon). Standard bending tests could not be done on this transitional tissue. Torsion testing of the soft collagen bridge did not seem to be sensitive enough. Using a specially designed apparatus for tensile testing, significant differences were measured at the time of bony bridging.15 After the distraction gap bridged with direct bone formation, remodeling to normal bone always progressed at standard rates. Therefore, early tensile testing of the distraction gap at the end of distraction gave significant insight as to the eventual success of later bone bridging, allowing for earlier sacrifice in future experiments.

The cell counting methods using NIH Image Analysis for immunohistochemistry were validated by comparing PCNA to H3 thymidine. In situ hybridization and immunohistochemistry for FGF-2 were consistent for mRNA and protein expression in cells having transition to angioblasts and osteoblasts.

Several delivery systems were used successfully in rats of different ages and having distraction osteogenesis. Some systems required staged surgeries whereas others could be done simultaneously. Preliminary studies confirmed local diffusion of inert markers into the distraction site, as other investigators have shown.36 Effective delivery of various substances was confirmed by serum testing and positive or negative modulation of bone formation. From this work, several potential models have emerged for the study of bone formation.

A well-rounded diet to avoid weight loss is important for adequate bone formation during distraction osteogenesis. Supplemental (supraphysiologic levels) sex or growth hormones do not accelerate the normal process of bone formation in young rats without an underlying deficiency. Acute calcium deficiency did not adversely affect the early process of bone formation in distraction osteogenesis in young rats if other dietary requirements were met. The early effects of ovariectomy had no effect on distraction osteogenesis in young rats. Lead exposure adversely affected distraction osteogenesis in addition to normal bone growth; these inhibitions were not reversed with administration of exogenous growth or sex hormones. Ethanol exposure also adversely affected distraction osteogenesis despite normal nutritional status. Different sources of protein in the diet, possibly associated with phytoestrogen isoflavones, may have beneficial effects on bone formation during distraction osteogenesis.

The deficits in bone formation measured in older rats strongly paralleled the radiographic findings of distraction osteogenesis when applied clinically in older patients.5,31 Selected biopsies of patients with poor mineralization of distraction gaps showed loose fibrous matrix without zones of new bone formation, similar to the histologic findings in old rats. The long delays in ossification across the distraction gap in adults, more than twice the time measured in children, are a critical problem facing surgeons who use distraction osteogenesis to regenerate bone deficiencies in adults. The extended time of external fixation leads to increasing prevalence and severity of complications including pin tract infections, muscle atrophy, muscle contractures, joint stiffness, and generalized disuse osteopenia.5,27,55

This aging model has certain other features that strongly resemble the much more prevalent disease of senile osteoporosis.43 Patients with senile osteoporosis experience primary loss of endosteal trabecular bone, marrow replacement of progenitor cells with fat, and increasing bone diameter with decreasing cortical thickness.33 The latter change in the macrostructure of bones results from the combined loss of endosteal trabecular bone and persistence of periosteal new bone formation which expands the external cortical diameter. Elderly patients are known to retain periosteal healing of fractures although the time of healing is delayed. Compression fractures of vertebrae and weakening of the femoral neck and the distal radius are related to loss of trabecular bone. The model for the study of endosteal bone formation in older rats not only allows one to stretch the process of direct bone formation into specific cellular zones, similar to a growth plate, but also allows measurement of the deficits during aging at a cellular level in vivo.

The distraction osteogenesis model has been scaled down to the rat with reliable zones of progressive intramembranous ossification, and the zonal transitions using molecular techniques to mechanical testing have been evaluated. A range of delivery systems and various modulators were used to test the potential effects on bone formation with this model. An aging model that is useful for testing the pathophysiologic features of diminished bone formation during aging and for testing potential pharmacologic treatments to restore normal bone formation was introduced. Specifically, the potential mechanisms of the FGF-2 axis of local control in coupled osteogenesis and angiogenesis during bone formation in this model were evaluated. Finally, the age-related deficits of endosteal bone formation were reversed by administering exogenous rhFGF-2. In doing so, the methods of administration were refined and a local dose effect was seen.

Future studies should be focused on cell-based therapy to address the observed marrow changes with aging43 and on the scaled down version of tibial lengthening in the mouse model, where transgenic mice51 are available to additionally test mechanisms and potential therapies for improving bone formation.

Back to Top | Article Outline

Acknowledgments

I thank my colleagues in research, Charles Lumpkin, Jr, Lichu Liu, Xinchu Shen, Daniel Perrien, Elizabeth Brown-Wahl, Thomas Badger, Robert Skinner, Ken Morris, Ron Tribble, William Hogue, Charlene Flahiff, Guan Gao, David Irby, Hyun-Dae Shin, Zhendong Liu, Seok Myun Ko, Jae-Young Rho, Martin Ronis, Francis Miller, and Tammy Quattlebaum, for their work.

Back to Top | Article Outline

References

1. Aronson J: Proper wire tensioning for Ilizarov external fixation. Tech Orthop 5:27–32, 1990.
2. Aronson J: The Biology of Distraction Osteogenesis. In Bianchi Maiocchi A, Aronson J (eds). Operative Principles of Ilizarov. Fracture Treatment, Nonunion, Osteomyelitis, Lengthening, Deformity Correction. Baltimore, Williams and Wilkins 42–52, 1991.
3. Aronson J: Cavitary osteomyelitis treated by fragmentary cortical bone transportation. Clin Orthop 280:153–159, 1992.
4. Aronson J: Temporal and spatial increases in blood flow during distraction osteogenesis. Clin Orthop 301:124–131, 1994.
5. Aronson J: Experimental and clinical experience with distraction osteogenesis. Cleft Palate Craniofac J 31:473–481, 1994.
6. Aronson J: Experimental Assessment of Bone Regenerate Quality During Distraction Osteogenesis. In Brighton CT, Friedlaender G, Lane JM (eds). Bone, Formation and Repair. Rosemont IL, American Academy of Orthopaedic Surgeons 441-463, 1994.
7. Aronson J: Limb-lengthening, skeletal reconstruction and bone transport with the Ilizarov method. J Bone Joint Surg 79A:1243–1258, 1997.
8. Aronson J, Gao GG, Shen XC, et al: The effect of aging on distraction osteogenesis in the rat. J Orthop Res 19:421–427, 2001.
9. Aronson J, Good B, Stewart C, Harrison B, Harp J: Preliminary studies of mineralization during distraction osteogenesis. Clin Orthop 250:43–49, 1990.
10. Aronson J, Harp JH: Mechanical forces as predictors of healing during tibial lengthening by distraction osteogenesis. Clin Orthop 301:73–79, 1994.
11. Aronson J, Harp Jr JH: Factors influencing the choice of external fixation for distraction osteogenesis. Instr Course Lect 39:175–183, 1990.
12. Aronson J, Harp Jr JH: Mechanical considerations in using tensioned wires in a transosseous external fixation system. Clin Orthop 280:23–29, 1992.
13. Aronson J, Harrison B, Boyd CM, Cannon DJ, Lubansky HJ: Mechanical induction of osteogenesis: The importance of pin rigidity. J Pediatr Orthop 8:396–401, 1988.
14. Aronson J, Harrison BH, Stewart CL, Harp Jr JH: The histology of distraction osteogenesis using different external fixators. Clin Orthop 241:106–116, 1989.
15. Aronson J, Hogue WR, Flahiff CM, et al: Development of tensile strength during distraction osteogenesis in a rat model. J Orthop Res 19:64–69, 2001.
16. Aronson J, Johnson E, Harp JH: Local bone transportation for treatment of intercalary defects by the Ilizarov technique: Biomechanical and clinical considerations. Clin Orthop 243:71–79, 1989.
17. Aronson J, Shen X: Experimental healing of distraction osteogenesis comparing metaphyseal with diaphyseal sites. Clin Orthop 301:25–30, 1994.
18. Aronson J, Shen XC, Gao GG, et al: Sustained proliferation accompanies distraction osteogenesis in the rat. J Orthop Res 15:563–569, 1997.
19. Aronson J, Shen XC, Skinner RA, et al: Rat model of distraction osteogenesis. J Orthop Res 15:221–226, 1997.
20. Badger TM, Crouch J, Irby D, Hakkak R, Shahare M: Episodic excretion of ethanol during chronic intragastric ethanol infusion in the male rat: Continuous vs. cyclic ethanol and nutrient infusions. J Pharmacol Exp Ther 264:938–943, 1993.
21. Badger TM, Millard WJ, Owens SM, LaRovere J, O’Sullivan D: Effects of gonadal steroids on clearance of growth hormone at steady state in the rat. Endocrinology 128:1065–1072, 1991.
22. Badger TM, Ronis MJ, Lumpkin CK, et al: Effects of chronic ethanol on growth hormone secretion and hepatic cytochrome P450 isozymes of the rat. J Pharmacol Exp Ther 264:438–447, 1993.
23. Brighton CT, Lorich DG, Kupcha R, et al: The pericyte as a possible osteoblast progenitor cell. Clin Orthop 275:287–299, 1992.
24. Bryant SR, Bjercke RJ, Erichsen D, Rege A, Lindner V: Vascular remodeling in response to altered blood flow is mediated by fibroblast growth factor-2. Circ Res 84:323–328, 1999.
25. Choi IH, Ahn JH, Chung CY, Cho TJ: Vascular proliferation and blood supply during distraction osteogenesis: A scanning electron microscopic observation. J Orthop Res 18:698–705, 2000.
26. Crabb ID, Hughes SS, Hicks DG, et al: Nonradioactive in situ hybridization using digoxigenin-labeled oligonucleotides: Applications to musculoskeletal tissues. Am J Pathol 141:579–589, 1992.
27. Dahl MT, Gulli B, Berg T: Complications of limb lengthening: A learning curve. Clin Orthop 301:10–18, 1994.
28. Delloye C, Delefortrie G, Coutelier L, Vincent A: Bone regenerate formation in cortical bone during distraction lengthening: An experimental study. Clin Orthop 250:34–42, 1990.
29. Dyer SA, Buckendahl P, Sampson HW: Alcohol consumption inhibits osteoblastic cell proliferation and activity in vivo. Alcohol 16:337–341, 1998.
30. Eyres KS, Bell MJ, Kanis JA: New bone formation during leg lengthening: Evaluated by dual energy x-ray absorptiometry. J Bone Joint Surg 75B:96–106, 1993.
31. Fischgrund J, Paley D, Suter C: Variables affecting time to bone healing during limb lengthening. Clin Orthop 301:31–37, 1994.
32. Frierson M, Ibrahim K, Boles M, Bote H, Ganey T: Distraction osteogenesis: A comparison of corticotomy techniques. Clin Orthop 301:19–24, 1994.
33. Fujita T: Global assessment of risk factors for osteoporosis. J Bone Miner Metab 19:131–132, 2001.
34. Ganey TM, Klotch DW, Sasse J, Ogden JA, Garcia T: Basement membrane of blood vessels during distraction osteogenesis. Clin Orthop 301:132–138, 1994.
35. Goldstein SA, Waanders N, Guldberg R, et al: Stress Morphology Relationships During Distraction Osteogenesis: Linkages Between Mechanical and Architectural Factors in Molecular Regulation. In Brighton CT, Friedlaender G, Lane JM (eds). Bone Formation and Repair. Rosemont IL, American Academy of Orthopaedic Surgeons 405–419, 1994.
36. Grayson BH, Rowe NM, Hollier LH, et al: Development of a device for the delivery of agents to bone during distraction osteogenesis. J Craniofac Surg 12:19–25, 2001.
37. Harp JH, Aronson J, Hollis M: Noninvasive determination of bone stiffness for distraction osteogenesis by quantitative computed tomography scans. Clin Orthop 301:42–48, 1994.
38. Ilizarov GA: The tension-stress effect on the genesis and growth of tissues: Part II. The influence of the rate and frequency of distraction. Clin Orthop 239:263–285, 1989.
39. Ilizarov GA: Clinical application of the tension-stress effect for limb lengthening. Clin Orthop 250:8–26, 1990.
40. Ilizarov GA: The tension-stress effect on the genesis and growth of tissues: Part I. The influence of stability of fixation and soft-tissue preservation. Clin Orthop 238:249–281, 1989.
41. Ippolito E, Peretti G, Bellocci M, et al: Histology and ultrastructure of arteries, veins, and peripheral nerves during limb lengthening. Clin Orthop 308:54–62, 1994.
42. Jazrawi LM, Majeska RJ, Klein ML, et al: Bone and cartilage formation in an experimental model of distraction osteogenesis. J Orthop Trauma 12:111–116, 1998.
43. Kahn A, Gibbons R, Perkins S, et al: Age-related bone loss: A hypothesis and initial assessment in mice. Clin Orthop 313:69–75, 1995.
44. Karp NS, McCarthy JG, Schreiber JS, Sissons HA, Thorne CH: Membranous bone lengthening: A serial histological study. Ann Plast Surg 29:2–7, 1992.
45. Kawaguchi H, Kurokawa T, Hanada K, et al: Stimulation of fracture repair by recombinant human basic fibroblast growth factor in normal and streptozotocin-diabetic rats. Endocrinology 135:774–781, 1994.
46. Klein RF: Alcohol-induced bone disease: Impact of ethanol on osteoblast proliferation. Alcohol Clin Exp Res 21:392–399, 1997.
47. Lieber CS, DeCarli LM: The feeding of alcohol in liquid diets: Two decades of applications and 1982 update. Alcohol Clin Exp Res 6:523–531, 1982.
48. Liu Z, Luyten FP, Lammens J, Dequeker J: Molecular signaling in bone fracture healing and distraction osteogenesis. Histol Histopathol 14:587–595, 1999.
49. Lumpkin Jr CK, Aronson J, Shen XC, et al: The impact of total enteral nutrition on distraction osteogenesis in a rat model. J Bone Miner Res 11:962–969, 1996.
50. Mehrara BJ, Rowe NM, Steinbrech DS, et al: Rat mandibular distraction osteogenesis: II. Molecular analysis of transforming growth factor beta-1 and osteocalcin gene expression. Plast Reconstr Surg 103:536–547, 1999.
51. Montero A, Okada Y, Tomita M, et al: Disruption of the Fibroblast Growth Factor-2 gene in mice results in decreased. bone mass and bone formation. J Clin Invest 105:1085–1093, 2000.
52. Montone KT, Brigati DJ: In situ molecular pathology: Instrumentation, oligonucleotides and viral nucleic acid detection. J Histotechnol 17:195–201, 1994.
53. Nakamura T, Hanada K, Tamura M, et al: Stimulation of endosteal bone formation by systemic injections of recombinant fibroblast growth factor in rats. Endocrinology 136:1276–1284, 1995.
54. Paley D: Current techniques of limb lengthening. J Pediatr Orthop 8:73–92, 1988.
55. Paley D: Problems, obstacles, and complications of limb lengthening by the Ilizarov technique. Clin Orthop 250:81–104, 1990.
56. Radomsky ML, Aufdemorte TB, Swain LD, et al: Novel formulation of FGF-2 in a hyaluronan gel accelerates fracture healing in nonhuman primates. J Orthop Res 17:607–614, 1999.
57. Ronis MJ, Aronson J, Gao GG, et al: Skeletal effects of developmental lead exposure in rats. Toxicol Sci 62:321–329, 2001.
58. Ronis MJ, Lumpkin CK, Ingelman-Sundberg M, Badger TM: Effects of short-term ethanol and nutrition on the hepatic microsomal monooxygenase system in a model utilizing total enteral nutrition in the rat. Alcohol Clin Exp Res 15:693–699, 1991.
59. Rowe NM, Mehrara BJ, Luchs JS, et al: Angiogenesis during mandibular distraction osteogenesis. Ann Plast Surg 42:470–475, 1999.
60. Schenk RK, Gachter A: Histology of Distraction Osteogenesis. In Brighton CT, Friedlaender G, Lane JM (eds). Bone Formation and Repair. Rosemont IL, American Academy of Orthopaedic Surgeons 387-394, 1994.
61. Seghezzi G, Patel S, Ren CJ, et al: Fibroblast growth factor (FGF-2) induces vascular endothelial growth factor expression in the endothelial cells of forming capillaries: An autocrine mechanism contributing to angiogenesis. J Cell Biol 141:1659–1673, 1998.
62. Shearer JR, Roach HI, Parsons SW: Histology of a lengthened human tibia. J Bone Joint Surg 74B:39–44, 1992.
63. Skinner RA, Hickmon SG, Lumpkin Jr CK, Aronson J, Nicholas RW: Decalcified bone: Twenty years of successful specimen managment. Histotechnol 16:765–773, 1997.
64. Trueta J: The role of the vessels in osteogenesis. J Bone Joint Surg 45B:402–418, 1963.
65. Tsukamoto H, Reidelberger RD, French SW, Largman C: Long-term cannulation model for blood sampling and intragastric infusion in the rat. Am J Physiol 247:R595–R599, 1984.
66. Vauhkonen M, Peltonen J, Karaharju E, Aalto K, Alitalo I: Collagen synthesis and mineralization in the early phase of distraction bone healing. Bone Miner 10:171–181, 1990.
67. Walker CW, Aronson J, Kaplan PA, Molpus WM, Seibert JJ: Radiologic evaluation of limb-lengthening procedures. Am J Roentgenol 156:353–358, 1991.
68. Walsh WR, Hamdy RC, Ehrlich MG: Biomechanical and physical properties of lengthened bone in a canine model. Clin Orthop 306:230–238, 1994.
69. Wolfson N, Hearn TC, Thomason JJ, Armstrong PF: Force and stiffness changes during Ilizarov leg lengthening. Clin Orthop 250:58–60, 1990.
70. Yasui N, Kojimoto H, Shimizu H, Shimomura Y: The effect of distraction upon bone, muscle, and periosteum. Orthop Clin North Am 22:563–567, 1991.
71. Yeung HY, Lee SK, Fung KP, Leung KS: Expression of basic fibroblast growth factor during distraction osteogenesis. Clin Orthop 385:219–229, 2001.
72. Ziotis J, Malamitsi-Puchner A, Tsonou A, et al: Basic fibroblast growth factor: Serum levels in the female. Growth Factors 17:215–220, 2000.
© 2004 Lippincott Williams & Wilkins, Inc.